FDA Removes Significant Barrier to Using Real-World Evidence in Product Submissions

The U.S. Food and Drug Administration announced a major shift in its approach to real‑world evidence (RWE). In new guidance for certain medical device submissions, FDA will accept RWE without always requiring identifiable, patient‑level data collected from real‑world data (RWD) sources to be submitted in the marketing application. FDA also signaled its intent to consider updating guidance for drugs and biologics to reflect a similar policy. Reviewers will evaluate the strength and suitability of RWE case‑by‑case.

What the policy does

• FDA will consider RWE drawn from de‑identified RWD without mandatory submission of identifiable patient‑level records if the evidence is fit‑for‑purpose for the regulatory question.
• Large de‑identified registries and multi‑institutional datasets can serve as primary or supportive evidence where methods, data quality, and provenance are robust.

Why it matters

• This development meaningfully expands the usability of de‑identified and privacy‑preserving datasets—such as national registries, integrated delivery network EHRs, insurance claims, and multi‑institutional data collaboratives—for regulatory decision‑making.
• It also responds to long‑standing sponsor concerns that mandatory patient‑level data submissions made many high‑value data sources effectively unusable, despite their scale and representativeness.

Immediate implications

• Medical devices: Guidance is effective for certain submissions. Sponsors can leverage de‑identified registries, claims, and EHR networks to support 510(k)s, PMAs, labeling changes, post‑market studies, and surveillance. Valid scientific evidence remains the standard; document RWD relevance, analytic rigor, endpoint validity, and confounding variable control. Be ready to explain data lineage and reproducibility without sharing identifiable records.
• Drugs and biologics: Although the device guidance changes first, the Agency’s intent to consider analogous updates for drugs and biologics is clear. Proceed under current RWE guidance while designing non‑interventional studies and externally controlled comparisons that use de‑identified sources. Substantial evidence remains the legal standard. Maintain transparency, auditability, and arrangements for FDA access to underlying sources if requested.

Where this is most impactful

• Oncology and rare diseases, where robust registries and longitudinal data exist, and randomized controls may be impracticable or unethical.
• Label expansions for therapies with established safety/benefit profiles, where RWE can efficiently contextualize effectiveness in routine practice.
• Post‑market commitments and performance surveillance, where de‑identified, continuously refreshed datasets can generate timely insights.

Action checklist

• Reassess near‑term submissions for which de‑identified RWE could now play a primary or corroborative role.
• Prioritize contracts with registries and networked data partners that enable FDA‑acceptable access without sharing identifiable records.
• Update internal RWE SOPs to codify transparency, provenance, and reproducibility standards aligned with FDA expectations.
• Refresh risk registers to address residual bias, data linkage gaps, and verification contingencies.
• Plan and schedule early FDA interactions to validate your RWE strategy and documentation package.

Sponsors that pivot quickly to de‑identified, fit‑for‑purpose data and rigorous methods can capitalize on this shift and accelerate development timelines while maintaining evidentiary integrity.

If you have any questions, please contact Georgia Ravitz, Shari Esfahani, Thora Johnson or another Orrick team member.